Fig. 2
From: How to approach machine learning-based prediction of drug/compound–target interactions
t-SNE based visualization of conventional (apaac, k-sep_pssm, pfam, taap) and learned (protvec, seqvec, transformer-avg, unirep1900) protein representations on; (a) enzymes including hydrolases, oxidoreductases, proteases, transferases, and other-enzymes groups, and (b) non-enzyme protein families including epigenetic regulators, ion channels, membrane receptors, transcription factors, and transporters, in different colors